286 ALASKAN GLACIER STUDIES with them in believing that vegetation does not gain a foothold almost at once on areas uncovered by glacial ice where unconsolidated deposits are present and seeding possible. They have estimated that this advance took place 60 years or so ago, while we are more inclined to ascribe the building, or at least the last modification of this moraine, and the destruction of these trees, to the advance that occurred about 1892 or to an earlier advance only a few years before that date, and certainly not as long ago as 1859. A section revealed in this glacial deposit by a wave-cut cliff near the western end shows the following succession, indicating at least two advances here. SECTION OF GLACIAL DRIFT NEAR COLUMBIA GLACIER Present surface, with morainic topography and growing shrubs. Till, with blue clay and angular, striated stones. Trunks of trees, and cones and fragments of bark. Till, with large angular glaciated bowlders. Bedrock, with glacial striae. The lower till surface was an undulating one, evidently produced by irregular erosion of the lower till mass. Former Stages of Columbia Glacier. There are many facts to prove that Columbia Glacier is now far less extensive than formerly, and that at an earlier stage it was a tributary of the great Prince William Sound glacier. Near Columbia Glacier the cirques, hanging valleys, roches moutonn^es hills, over-steepened lower slopes, truncated spurs, the generally rounded topography produced by glacial sculpture, and the glacial deposits testify to this former expanded stage of glaciation. How thick the ice was at maximum has not been determined exactly. Heather Island, 857 feet high, was glaciated to the top. The mountain spur southeast of the ice tongue was glaciated as high as we climbed it, the elevation of the photographic station at timber line there being 1497 feet. From a distance the great nunatak has the rounded appearance of having been glaciated clear to the top, 3664 feet, though this is not certain. Gilbert places the height of glaciation in Columbia Bay at 4000 feet.1 Between this great maximum and the present stage of expansion, and with unknown intermediate episodes, there has evidently been notable recession of the glacier, proved by forests upon the mountain spurs above the upper glacier for a distance of at least six or seven miles from the present terminus, and upon the lower slopes of the great nunatak. It is conceivable, though hardly probable, that the forest may have migrated to the nunatak along the eastern ice margin with the glacier as expanded as now; but it is much more difficult to believe that the forest on the western side has migrated so far up the margin of a clear ice glacier. The farthest trees seen were on the northern side of the second embayment, or tributary, on the western side of the glacier. They seemed to be spruces. The greenish tint of the lower slopes beyond the next tributary suggested vegetation there, perhaps shrubs like the alder, but no trees could be detected with the field glasses. This last locality is over ten miles from the present ice front. Gilbert interpreted the bending of moraines into the first embayment as evidence that the 1892 maximum was preceded by an important minimum,2 and we are inclined to interpret this distribution of forest in the same way. Grant and Higgins have inferred a retreat i Op. dt., p. 174. « Op. cit., p. 80.